Structural cereal model

import warnings
warnings.filterwarnings('ignore', category=DeprecationWarning)

In this section 

3.1 Create a parametric leaf

• x,y coordinates of points sampling a leaf midrib placed in a vertical plane (origin = leaf base)
• s,r parallel array for curviliear abcissa / relative leaf width along leaf shape

import numpy as np
import matplotlib.pyplot as plt

from heapq import *
from scipy.interpolate import splprep, splev
from scipy.integrate import simps, trapz
from openalea.plantgl.all import Vector3

from cereals_leaf import leaf_shape_perez, sr_prevot, parametric_leaf
# or
# from simple_maize import leaf_shape_perez, sr_prevot, parametric_leaf
# from fitting import leaf_shape_perez

from generator import curvilinear_abscisse

from fitting import fit2, fit3, simplify

from simplification import distance, cost


pl=parametric_leaf(nb_segment=10, insertion_angle=40, scurv=0.7, curvature=70, alpha=-2.3)

fig, (ax1, ax2) = plt.subplots(nrows=2) 
fig.suptitle('Parametric leaf')

ax1.plot(pl[0], pl[1], '.-', c="green")
ax1.set_xlabel('x')
ax1.set_ylabel('y')
ax1.set_title("Coordinates of the midrib")

ax2.plot(pl[2], pl[3], '.-', c="green")
ax2.plot(pl[2], -pl[3], '.-', c="green")
ax2.plot(np.arange(0,1.1,0.1), np.zeros(11), c="green", ls="dashed")
ax2.set_xlabel('s')
ax2.set_ylabel('r')
ax2.set_title("Leaf shape")

plt.subplots_adjust(hspace=0.5)

plt.show()

# Enable interactive plotting
%matplotlib ipympl

from mpl_toolkits.mplot3d import Axes3D
from scipy.interpolate import interp2d
import matplotlib.tri as mtri

from fitting import leaf_to_mesh_2d

xy=pl[0:2]
r=pl[3]

pts,ind=leaf_to_mesh_2d(xy[0], xy[1], r)

xs=[pt[0] for pt in pts]
ys=[pt[1] for pt in pts]
zs=[pt[2] for pt in pts]
    
X,Y=np.meshgrid(xs, ys)

tri = mtri.Triangulation(xs, ys)

fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
ax.scatter(xs,ys,zs,c="green")
ax.plot(xs,ys,zs,c="green")
ax.plot(xs,np.zeros(len(ys)),zs,c="green",ls="dashed")
ax.set_title("3D representation of a leaf shape")
plt.show()

3.2 Generate leaf azimuth series

# from itertools import cycle

from plant_design import leaf_azimuth

nb_phy=10
la=leaf_azimuth(size=nb_phy, phyllotactic_angle=137, phyllotactic_deviation=0, plant_orientation=0, spiral=True)

x = np.cos(la*np.pi/180)
y = np.sin(la*np.pi/180)
z = np.linspace(1,len(la)+1,len(la))

fig, ax = plt.subplots(subplot_kw=dict(projection='3d'))
# ax.scatter(x, y, z, c="green")
for i,a in enumerate(la):
    ax.plot(np.linspace(0,x[i],2), np.linspace(0,y[i],2), [z[i],z[i]], c="green")
ax.plot([0,0], [0,0], [0,z[-1]], c="green")

ax.set_title("3D representation of phyllotaxy")

plt.show()

3.3 Manage internode lengths

def geometric_dist(height=15, nb_phy=15, q=1):
    """ returns distances between individual leaves along a geometric model """

    if q == 1:
        u0 = float(height) / nb_phy
    else:
        u0 = height * (1. - q) / (1. - q ** (nb_phy + 1))

    return [u0 * q ** i for i in range(nb_phy)]

x = np.cos(la*np.pi/180)
y = np.sin(la*np.pi/180)
z = geometric_dist(height=15, nb_phy=nb_phy, q=1.5)

fig, ax = plt.subplots(subplot_kw=dict(projection='3d'))
for i,a in enumerate(la):
    ax.plot(np.linspace(0,x[i],2), np.linspace(0,y[i],2), [z[i],z[i]], c="green")
ax.plot([0,0], [0,0], [0,z[-1]], c="green")

ax.set_title("3D representation of the repartition of internode length along the stem")

plt.show()

3.4 Manage leaf lengths as a function of height

def bell_shaped_dist(plant_area=1, nb_phy=15, rmax=.7, skew=0.15):
    """ returns leaf area of individual leaves along bell shaped model """

    k = -np.log(skew) * rmax
    r = np.linspace(1. / nb_phy, 1, nb_phy)
    relative_surface = np.exp(-k / rmax * (2 * (r - rmax) ** 2 + (r - rmax) ** 3))
    leaf_area = relative_surface / relative_surface.sum() * plant_area
    return leaf_area.tolist()

bsd=bell_shaped_dist(plant_area=1, nb_phy=nb_phy, rmax=.7, skew=0.15)
x = np.cos(la*np.pi/180)*bsd
y = np.sin(la*np.pi/180)*bsd
z = geometric_dist(height=15, nb_phy=nb_phy, q=1.5)

fig, ax = plt.subplots(subplot_kw=dict(projection='3d'))
for i,a in enumerate(la):
    ax.plot(np.linspace(0,x[i],2), np.linspace(0,y[i],2), [z[i],z[i]], c="green")
ax.plot([0,0], [0,0], [0,z[-1]], c="green")

ax.set_title("3D representation of the repartition of leaf length along the stem")

plt.show()

3.5 Arrange a leaf to be placed along a stem with a given inclination.

from math import pi, cos, sin, radians
import openalea.plantgl.all as pgl

# from cereals_leaf import arrange_leaf
# or 
from geometry import arrange_leaf

stem_diameter=0.5
inclination=1.2

al=arrange_leaf(leaf=pl, stem_diameter=stem_diameter, inclination=inclination, relative=True)


x=al[0]
y=al[1]
s=al[2]
r=al[3]

pts,ind=leaf_to_mesh_2d(x, y, r)

xs=[pt[0] for pt in pts]
ys=[pt[1] for pt in pts]
zs=[pt[2] for pt in pts]
    
X,Y=np.meshgrid(xs, ys)

tri = mtri.Triangulation(xs, ys)

fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
# ax.scatter(xs,ys,zs,c="green")
ax.plot(xs,ys,zs,c="green")
ax.plot([xs[0],xs[0]],[ys[0],-ys[0]],[0,0],c="green")
ax.plot(xs,np.zeros(len(ys)),zs,c="green",ls="dashed")

radius=stem_diameter/2
z = np.linspace(0, zs[-1])
theta = np.linspace(0, 2*np.pi)
theta_grid, z_stem=np.meshgrid(theta, z)
x_stem = radius*np.cos(theta_grid)
y_stem = radius*np.sin(theta_grid)

ax.plot_surface(x_stem, y_stem, z_stem, color="green")

ax.set_title("3D representation of the placement of a leaf along a stem")

plt.show()

3.6 Build the whole plant shoot in 3D, as an MTG.

import openalea.plantgl.all as pgl
from openalea.mtg.turtle import TurtleFrame

from geometry import slim_cylinder, stem_mesh, _is_iterable, as_tuples, addSets, leaf_mesh, compute_element
# from cereals_leaf import leaf_mesh
from geometry import CerealsTurtle, CerealsVisitor

from fitting import leaf_element, leaf_to_mesh_2d, leaf_to_mesh, mesh4, plantgl_shape   # leaf_to_mesh_new ?

from openalea.mtg import MTG, fat_mtg
from scipy.interpolate import interp1d
import pandas

from plant_design import get_form_factor, blade_dimension, stem_dimension

from generator import majors_axes_regression, line_projection, as_leaf, as_plant, cereals as cereals_generator

from geometry import mtg_interpreter


def build_shoot(stem_radius, insertion_heights, leaf_lengths, leaf_areas, leaf_azimuths=None, leaf_shapes=None):
    """create a shoot

    Args:
        stem_radius: (float) the stem radius
        insertion_heights: list of each leaf insertion height
        leaf_lengths: list of each leaf length (blade length)
        leaf_areas: list of each blade area
        collar_visible: list of each collar height or True if the collar is visible and False if it is not
        leaf_shapes: list of each leaf shape, if it is not known write None
        leaf_azimuths: list of each leaf azimuth, if it is not known write None

    Returns:
        shoot: 

    """
    ranks = range(1, len(leaf_lengths) + 1)
    ntop = max(ranks) - np.array(ranks) + 1
    if leaf_shapes is None:
        a_leaf = parametric_leaf()
        leaf_shapes = [a_leaf for r in ranks]
    if leaf_azimuths is None:
        leaf_azimuths = leaf_azimuth(len(ranks))
    leaf_azimuths[1:] = np.diff(leaf_azimuths)
    ff = [get_form_factor(leaf) for leaf in leaf_shapes]
    blades = blade_dimension(area=leaf_areas, length=leaf_lengths, ntop=ntop)
    stem = stem_dimension(h_ins=insertion_heights, d_internode=np.array(stem_radius) * 2, ntop=ntop)
    df = blades.merge(stem)
    df['leaf_azimuth'] = leaf_azimuths
    df['leaf_rank'] = ranks
    df['leaf_shape'] = [leaf_shapes[n - 1] for n in df.leaf_rank]
    return df, cereals_generator(plant=df)

def build_shoot_w_pseudo(nb_phy, plant_height, insertion_heights, leaf_lengths, leaf_areas, 
                pseudostem_dist=1.4, stem_dist=1.2,
                diam_base=2.5, diam_top=1, pseudostem_height=20, 
                leaf_azimuths=None, leaf_shapes=None, wl=0.1):
    """create a shoot, with pseudostems and stems

    Args:
        stem_radius: (float) the stem radius
        insertion_heights: list of each leaf insertion height
        leaf_lengths: list of each leaf length (blade length)
        leaf_areas: list of each blade area
        collar_visible: list of each collar height or True if the collar is visible and False if it is not
        leaf_shapes: list of each leaf shape, if it is not known write None
        leaf_azimuths: list of each leaf azimuth, if it is not known write None

    Returns:
        shoot: 

    """
    ranks = range(1, len(leaf_lengths) + 1)
    ntop = max(ranks) - np.array(ranks) + 1

    nb_phy = int(nb_phy)

    # Lejeune an Bernier formula + col =
    nb_young_phy = int(round((nb_phy - 1.95) / 1.84 / 1.3))

    # distances between leaves
    pseudostem = geometric_dist(pseudostem_height, nb_young_phy,
                                pseudostem_dist)
    stem = geometric_dist(plant_height - pseudostem_height,
                          nb_phy - nb_young_phy, stem_dist)
    internode = pseudostem + stem
    # stem diameters
    diameter = ([diam_base] * nb_young_phy +
                np.linspace(diam_base, diam_top,
                            nb_phy - nb_young_phy).tolist())
    
    if leaf_shapes is None:
        a_leaf = parametric_leaf()
        leaf_shapes = [a_leaf for r in ranks]
    if leaf_azimuths is None:
        leaf_azimuths = leaf_azimuth(len(ranks))
    leaf_azimuths[1:] = np.diff(leaf_azimuths)
    ff = [get_form_factor(leaf) for leaf in leaf_shapes]
    # blades = blade_dimension(area=leaf_areas, length=leaf_lengths, ntop=ntop)
    blades = blade_dimension(length=leaf_lengths, form_factor=ff, ntop=ntop, wl=wl)
    # stem = stem_dimension(h_ins=insertion_heights, d_internode=diameter, ntop=ntop)
    stem = stem_dimension(internode=internode, d_internode=diameter, ntop=ntop)
    
    df = blades.merge(stem)
    df['leaf_azimuth'] = leaf_azimuths
    df['leaf_rank'] = ranks
    df['leaf_shape'] = [leaf_shapes[n - 1] for n in df.leaf_rank]
    return df, cereals_generator(plant=df)

# shoot, g = build_shoot(3.0, [2,4,6,8], [2,4,6,8], [2,4,6,8])
# scene, nump = build_scene(g)
# display_scene(scene)

3.7 Display scenes according to different scenarii

3.7.1 A single cereal

from display import display_mtg, build_scene, display_scene
from oawidgets.plantgl import *
# from oawidgets.mtg import *

%gui qt

# generation of a 3D plant from descritive parameters
stem_radius=1 # realistic, number or growing with phytomer age ?
height=1500 # from crop model
nb_phy=15 # from 0 to fixed max nb of phytomers
max_leaf_length=800
insertion_angle=40
scurv=0.7
curvature=80
phyllotactic_angle=120
spiral=True

insertion_heights=np.array(geometric_dist(height, nb_phy, q=1.2)) # further separate stem and pseudo stem, cf simple maize

leaf_lengths=np.array(bell_shaped_dist(plant_area=max_leaf_length, nb_phy=nb_phy, rmax=0.7, skew=0.15)) # plant area --> max leaf length
# leaf_areas=bell_shaped_dist(plant_area=1, nb_phy=15, rmax=0.7, skew=0.15) # cf blade_dimension

a_leaf = parametric_leaf(nb_segment=10, insertion_angle=insertion_angle, scurv=scurv, curvature=curvature, alpha=-2.3)
leaf_shapes = [a_leaf for l in leaf_lengths] # replace leaf_length by nb_phy or...

leaf_azimuths = leaf_azimuth(size=len(leaf_lengths), phyllotactic_angle=phyllotactic_angle, phyllotactic_deviation=15, plant_orientation=0, spiral=spiral)

# shoot, g = build_shoot_w_pseudo(nb_phy=nb_phy, plant_height=height, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None, 
#                 pseudostem_dist=1.4, stem_dist=1.,
#                 diam_base=2.5, diam_top=1, pseudostem_height=20, 
#                 leaf_azimuths=None, leaf_shapes=None)

shoot, g_single = build_shoot(stem_radius=stem_radius, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None,
                leaf_shapes=leaf_shapes, leaf_azimuths=leaf_azimuths)


# g_single.display()

scene_single, nump = build_scene(g_single, leaf_material=Material(Color3((50,100,0))), stem_material=Material(Color3((50,100,0))))
# display_scene(scene_single)
PlantGL(scene_single)

from oawidgets.mtg import *

# Properties on the MTG: this exclude all the topological properties
print(g_single.property_names())

# Retrieve one property for the MTG (dict)

labels = g_single.property('label')
# print(labels)

length = g_single.property('length')
# print(length)

leaf_lengths=[]
leaf_ind=[]
internode_lengths=[]
internode_ind=[]
for k,v in length.items():
    if k%2==0: # could have done it using labels
        internode_ind.append(k)
        internode_lengths.append(v)
    else:
        leaf_ind.append(k)
        leaf_lengths.append(v)

width = g_single.property('shape_max_width')
# print(width)

leaf_widths=[]
for k,v in width.items():
    leaf_widths.append(v)


plt.figure()
plt.plot(leaf_ind, leaf_lengths, c="green", label="Leaf lengths") # == 'shape_mature_length' for final plant
plt.plot(leaf_ind, leaf_widths, c="purple", label="Leaf widths")
plt.plot(internode_ind, internode_lengths, c="orange", label="Internode lengths")
plt.xlabel("Vertices")
plt.ylabel("Lengths (cm)")
plt.title("MTG properties")
plt.legend()
plt.show()

['edge_type', 'label', 'length', 'is_green', 'diameter_base', 'diameter_top', 'azimuth', 'shape', 'shape_mature_length', 'visible_length', 'srb', 'srt', 'lrolled', 'd_rolled', 'shape_max_width', 'stem_diameter', 'geometry', 'anchor_point', 'color']

The leaf lengths and widths follow the bell shaped curve described before.
The first internode in the MTG actually corresponds to the pseudostem, i.e. the about 4 to 8 short first internodes that rapidly lose their leaves.

# classes = list(set(g.class_name(vid) for vid in g.vertices() if g.class_name(vid)))
# print(classes)

# def vertices(g, class_name='P'):
#     return [vid for vid in g.vertices() if g.class_name(vid)==class_name]

# vids_U = vertices(g, 'U')

# plot(g_single, selection=vids_U)

# plot(g_single, selection=[vid for vid in g_single.vertices() if g_single.class_name(vid)=='LeafElement'])

3.7.2 A cereal crop with variability

from stand import agronomic_plot
from display import display_mtg, build_scene, display_scene
from random import *
from oawidgets.plantgl import *
# from oawidgets.mtg import *

%gui qt

seed(1)

# generation of a crop of 3D plants from descritive parameters with variability
plants_in_crop=[]
length_plot=10
width_plot=10
sowing_density=10
plant_density=2
inter_row=2
nplants, positions, domain, domain_area, unit = agronomic_plot(length_plot, width_plot, sowing_density, plant_density, inter_row, noise=0.1)

for n in range(nplants):
    stem_radius=1 # realistic, number or growing with phytomer age ?
    height=2000*(1+random()-0.5) # from crop model
    nb_phy=15 # from 0 to fixed max nb of phytomers
    max_leaf_length=2000*(1+random()-0.5)
    insertion_angle=50*(1+random()-0.5)
    scurv=0.5*(1+random()-0.5)
    curvature=50*(1+random()-0.5)
    phyllotactic_angle=120*(1+random()-0.5)
    spiral=True
    
    insertion_heights=np.array(geometric_dist(height, nb_phy, q=1.2)) # further separate stem and pseudo stem, cf simple maize
    
    leaf_lengths=np.array(bell_shaped_dist(plant_area=max_leaf_length, nb_phy=nb_phy, rmax=0.7, skew=0.15)) # plant area --> max leaf length
    # leaf_areas=bell_shaped_dist(plant_area=1, nb_phy=15, rmax=0.7, skew=0.15) # cf blade_dimension
    
    a_leaf = parametric_leaf(nb_segment=10, insertion_angle=insertion_angle, scurv=scurv, curvature=curvature, alpha=-2.3)
    leaf_shapes = [a_leaf for l in leaf_lengths] # replace leaf_length by nb_phy or...
    
    leaf_azimuths = leaf_azimuth(size=len(leaf_lengths), phyllotactic_angle=phyllotactic_angle, phyllotactic_deviation=15, plant_orientation=0, spiral=spiral)
    
    # shoot, g = build_shoot_w_pseudo(nb_phy=nb_phy, plant_height=height, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None, 
    #                 pseudostem_dist=1.4, stem_dist=1.,
    #                 diam_base=2.5, diam_top=1, pseudostem_height=20, 
    #                 leaf_azimuths=None, leaf_shapes=None)
    shoot, g_var = build_shoot(stem_radius=stem_radius, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None,
                    leaf_shapes=leaf_shapes, leaf_azimuths=leaf_azimuths)

    plants_in_crop.append(g_var)


scene_var, nump = build_scene(plants_in_crop, positions, leaf_material=Material(Color3((50,100,0))), stem_material=Material(Color3((50,100,0))))
# display_scene(scene_var)
PlantGL(scene_var)

3.7.3 A seemingly growing plant

from stand import agronomic_plot
from display import display_mtg, build_scene, display_scene
from random import *
from openalea.plantgl.all import *
from oawidgets.plantgl import *
# from oawidgets.mtg import *

%gui qt

seed(1)

# generation of ""growing"" 3D plants from descritive parameters
growing_plants=[]
length_plot=1
width_plot=1
sowing_density=10
plant_density=10
inter_row=0.75
# nplants, positions, domain, domain_area, unit = agronomic_plot(length_plot, width_plot, sowing_density, plant_density, inter_row, noise=0.1)
nplants=10
positions=[(x,0,0) for x in range(-500, 500, 100)]
final_height=2000
final_nb_phy=2*nplants
heights=geometric_dist(final_height, final_nb_phy, 1.2)
# insertion_heights=np.array(geometric_dist(height, nb_phy, q=1.2))

for n in range(1,nplants+1):
    stem_radius=2 # realistic, number or growing with phytomer age ?
    height=10*heights[2*n-1] # from crop model
    nb_phy=2*n # from 0 to fixed max nb of phytomers
    max_leaf_length=100*2*n
    insertion_angle=30
    scurv=0.7
    curvature=100
    phyllotactic_angle=137
    spiral=True
    
    insertion_heights=np.array(geometric_dist(height, nb_phy, q=1.2)) # further separate stem and pseudo stem, cf simple maize
    
    leaf_lengths=np.array(bell_shaped_dist(plant_area=max_leaf_length, nb_phy=nb_phy, rmax=0.7, skew=0.15)) # plant area --> max leaf length
    # leaf_areas=bell_shaped_dist(plant_area=1, nb_phy=15, rmax=0.7, skew=0.15) # cf blade_dimension
    # print(leaf_lengths)
    a_leaf = parametric_leaf(nb_segment=10, insertion_angle=insertion_angle, scurv=scurv, curvature=curvature, alpha=-2.3)
    leaf_shapes = [a_leaf for l in leaf_lengths] # replace leaf_length by nb_phy or...

    leaf_azimuths = leaf_azimuth(size=len(leaf_lengths), phyllotactic_angle=phyllotactic_angle, phyllotactic_deviation=15, plant_orientation=0, spiral=spiral)

    # shoot, g = build_shoot_w_pseudo(nb_phy=nb_phy, plant_height=height, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None, 
    #                 pseudostem_dist=1.4, stem_dist=1.,
    #                 diam_base=2.5, diam_top=1, pseudostem_height=20, 
    #                 leaf_azimuths=None, leaf_shapes=None)
    shoot, g_grow = build_shoot(stem_radius=stem_radius, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None,
                    leaf_shapes=leaf_shapes, leaf_azimuths=leaf_azimuths)

    # print(g_grow)

    growing_plants.append(g_grow)


scene_grow, nump = build_scene(growing_plants[2:], positions[2:], leaf_material=Material(Color3((50,100,0))), stem_material=Material(Color3((50,100,0))))
# display_scene(scene_grow)
PlantGL(scene_grow)

3.7.4 An intercrop organized in rows

from stand import agronomic_plot
from display import display_mtg, build_scene, display_scene
from random import *
from oawidgets.plantgl import *
# from oawidgets.mtg import *
from openalea.plantgl.all import Material, Color3, Shape, Scene, Viewer, Translated, AxisRotated

%gui qt

seed(1)

# generation of 3D intercropped plants from descritive parameters
# plants_in_crop=[]
# length_plot=1
# width_plot=1
# sowing_density=10
# plant_density=10
# inter_row=0.75
# nplants, positions, domain, domain_area, unit = agronomic_plot(length_plot, width_plot, sowing_density, plant_density, inter_row, noise=0.1)
# nplants=10

# final_height=2000
# final_nb_phy=2*nplants
# heights=geometric_dist(final_height, final_nb_phy, 1.2)
# insertion_heights=np.array(geometric_dist(height, nb_phy, q=1.2))
def plant(height, nb_phy, max_leaf_length, phyllotactic_angle, spiral):
    stem_radius=1 # realistic, number or growing with phytomer age ?
    insertion_angle=30
    scurv=0.7
    curvature=100
    
    insertion_heights=np.array(geometric_dist(height, nb_phy, q=1.2)) # further separate stem and pseudo stem, cf simple maize
    
    leaf_lengths=np.array(bell_shaped_dist(plant_area=max_leaf_length, nb_phy=nb_phy, rmax=0.7, skew=0.15)) # plant area --> max leaf length
    # leaf_areas=bell_shaped_dist(plant_area=1, nb_phy=15, rmax=0.7, skew=0.15) # cf blade_dimension
    
    a_leaf = parametric_leaf(nb_segment=10, insertion_angle=insertion_angle, scurv=scurv, curvature=curvature, alpha=-2.3)
    leaf_shapes = [a_leaf for l in leaf_lengths] # replace leaf_length by nb_phy or...
    
    leaf_azimuths = leaf_azimuth(size=len(leaf_lengths), phyllotactic_angle=phyllotactic_angle, phyllotactic_deviation=15, plant_orientation=0, spiral=spiral)

    # shoot, g = build_shoot_w_pseudo(nb_phy=nb_phy, plant_height=height, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None, 
    #                 pseudostem_dist=1.4, stem_dist=1.,
    #                 diam_base=2.5, diam_top=1, pseudostem_height=20, 
    #                 leaf_azimuths=None, leaf_shapes=None)
    shoot, g = build_shoot(stem_radius=stem_radius, insertion_heights=insertion_heights, leaf_lengths=leaf_lengths, leaf_areas=None,
                    leaf_shapes=leaf_shapes, leaf_azimuths=leaf_azimuths)

    return g


### Mixture organized in alternate rows

n_rows = 10
len_rows = 10

d_inter = 100
d_intra = 70

def plant_in_row(i):
     if i%(4*d_inter)==0 or i%(4*d_inter)==d_inter: return plant(height=2000, nb_phy=15, max_leaf_length=1200, phyllotactic_angle=137, spiral=True)
     else: return plant(height=1000, nb_phy=10, max_leaf_length=400, phyllotactic_angle=137, spiral=True)

# ms=[]

# for x in range(0, n_rows*d_inter, d_inter):
#     for y in range(0, len_rows*d_intra, d_intra):
#         g, col=plant_in_row(x)
#         ts=Translated((x,y,0), g)
#         sh=Shape(ts, col)

#         ms.append(sh)

# scene=Scene(ms)

# Viewer.display(scene)

plants_in_intercrop = [plant_in_row(x) for x in range(0, n_rows*d_inter, d_inter) for y in range(0, len_rows*d_intra, d_intra)]

positions=[(x,y,0) for x in range(0, n_rows*d_inter, d_inter) for y in range(0, len_rows*d_intra, d_intra)]

scene_ic, nump = build_scene(plants_in_intercrop, positions, leaf_material=Material(Color3((50,100,0))), stem_material=Material(Color3((50,100,0))))
# display_scene(scene_ic)
PlantGL(scene_ic)

3.8 Tillering / Branching

# Copy MTG single plant
g_branch = g_single.sub_mtg(g_single.root)

# g_branch.display()

# REARRANGE MTG WITH SCALES: PLANT, AXE, METAMER (with internode and leaf), ELEMENTS (StemElements and LeafElements)

# Add branch 

from openalea.mtg import MTG

g = MTG()

plant_id = g.add_component(g.root, label='P1')

# scale 1
v1 = g.add_child(plant_id, label='A1')
v2 = g.add_child(plant_id, label='A2')
v3 = g.add_child(plant_id, label='A3')
v4 = g.add_child(v1, label='I1')
v5 = g.add_child(v1, label='I2')

g.display()

from oawidgets.mtg import *
# plot(g)

MTG : nb_vertices=7, nb_scales=2
/P1         (id=1)                                          
    /A1         (id=2)                                      
        /I1         (id=5)                                  
        /I2         (id=6)                                  
    /A2         (id=3)                                      
    /A3         (id=4)                                      

3.9 Light interception with Caribu

from alinea.caribu.CaribuScene import CaribuScene
from alinea.caribu.data_samples import data_path

sky = str(data_path('Turtle16soc.light'))
# opts = map(str, [data_path('par.opt'), data_path('nir.opt')])

# complete set of files
cs = CaribuScene(scene=scene_var, light=sky) # opt=opts, ) #pattern=pattern)
raw,agg=cs.run(simplify=True)

# print(raw.keys())

scene,values = cs.plot(raw['Ei'],display=False)

v99 = np.percentile(values, 99)
nvalues=np.array(values)
nvalues[nvalues>v99]=v99
values = nvalues.tolist()

PlantGL(scene, group_by_color=False, property=values)

from alinea.caribu.CaribuScene import CaribuScene
from alinea.caribu.data_samples import data_path

sky = str(data_path('Turtle16soc.light'))
# opts = map(str, [data_path('par.opt'), data_path('nir.opt')])

# complete set of files
cs = CaribuScene(scene=scene_ic, light=sky) # opt=opts, ) #pattern=pattern)
raw,agg=cs.run(simplify=True)

scene,values = cs.plot(raw['Ei'],display=False)

v99 = np.percentile(values, 99)
nvalues=np.array(values)
nvalues[nvalues>v99]=v99
values = nvalues.tolist()

PlantGL(scene, group_by_color=False, property=values)

3.10 Next steps:

• Dynamic FSPM : appearance and elongation of phytomers (internodes and leaves)
• Constrains from crop model
• Adapt for different species (parameters for cereals, some functions for legumes)